Homicide is the most serious form of violence against women, with clearly different dynamics from homicide between men. Homicide is the leading cause of death for African-American women and the seventh leading cause of premature death for women overall in this country. Adolescent and adult women are most often killed by a husband, lover, or ex-husband or lover and physically abused women are at most risk to be killed by an intimate partner. An approximately equal number of women kill their current or former intimate partners each year as are killed by them, but wife abuse is the most frequent precursor of this form of homicide also. In order to prevent this form of violence, specific risk factors for intimate partner homicide need to be established. This proposal will build on the PI's prior research in the area with a case control study to investigate the relative risk of factors listed on the Danger Assessment, an instrument with some prior evidence of reliability and construct validity. Other risk factors suggested by both clinical and recent research evidence will also be tested in a multi city case control design. Police records of all homicides involving women over the age of 13 as victims will be individually examined in 7 geographically varied cities (Baltimore, Tampa, Kansas City, Missouri and Kansas, Seattle, Chicago, and Houston). Cases of women involved in intimate partner homicide (husband-wife, lover, estranged/divorced couple) will be examined for the identified risk factors. Family member or female friend survivors will be called for supplementary information. Female controls in an intimate relationship with the same age range, neighborhood and history of intimate partner violence (yes/no) will be selected by survey phone calls. Approximately 250 cases and 250 controls will be used for logistic regression analysis. In addition, 250 women who have been shot by an intimate partner ("almost" homicides) also will be compared with the 250 controls. A random subset of 30 of the attempted femicides will also be interviewed in depth to determine additional risk factors and to gain additional insight into context.